致力于打造开放共享的金融业知识平台， 中金点睛数字化投研平台诚邀您深度体验 ！ 中金点睛是集成中金研究分析师投研智慧的一站式数字化投研服务平台，依托中金研究 30 多个专业团队、全球市场视野、超 1800 支个股覆盖的 深度积淀，提供研究报告、会议活动、基本面数据库、研究框架等分析师研究成果，并结合大模型技术，致力于为客户提供高效、专业、准确的研 究服务。 点击图片，即刻登录体验 r store the true to the month to the sent to www.research.cicc.com），手机号登录即享 ■ 研究观点 日度更新投研焦点,精选文章及时推送 中金晨报 2 公开直播 资深分析师及时解读市场热点 公开路海 B 精品视频 用年下文 真人出镜,图文并茂,直观展示 CICC REITs TALK 2,8 认证即享 升级权益 邮箱认证,解锁三大升级功能 ■ 研究报告 3W+ 完整版研报 宏观经济 行业研究 大宗商品 � 数据与研究框架 160+ 行业研究框架 行业数据 40+ 精品数据库 精品数据看板 8 中金点晴大模型 - (1) 0 o AI搜索 1 要点筛理 智能问答 智能搜索 ...

Core Insights - The article highlights that Mianbi Intelligent has completed a new round of financing amounting to several hundred million yuan, aimed at enhancing the development of large models and accelerating commercialization efforts [1][4]. Company Overview - Mianbi Intelligent was established in 2022, focusing on innovation and application transformation in large model technology [4]. - The founding team is from Tsinghua University, specifically the Natural Language Processing Laboratory [4]. - Li Dahai, previously with Zhihu, became the CEO of Mianbi Intelligent in 2023, while Liu Zhiyuan, a co-founder and chief scientist, is an associate professor in the Department of Computer Science and Technology at Tsinghua University [4]. Financing Details - The recent financing round was led by Beijing's state-owned investment platform "Jingguorui" and market-oriented venture capital fund "Mijuhua" [1]. - The funds raised will primarily be used to increase research and development efforts on the end-side large models and to promote the commercialization process [1].

驱车沿徐东大街向长江二桥方向行驶，一座座高楼直插云霄，刷新着城市天际线。 在徐东大街与临江大道两侧，阿里巴巴、国铁城投等一批龙头企业纷纷落子，沿街的18栋楼宇已集聚1900余家企业，其中数字经 济相关企业接近1000家。这片约5.6平方公里的区域，正在武汉的产业版图上勾勒出一条宽阔的产业"走廊"。 今年以来平均每月新增50家企业 "今年以来，这里平均每月新增50家企业，已形成武昌区最核心、最具活力的数创街区。"武汉滨江数创走廊管委会工作专班负责 人贺耀辉介绍。 武汉数创大厦29层，一位年轻工程师手握类似游戏手柄的物件，轻轻转动手臂，眼前的人形机器人机械臂"神同步"划出一道弧 线。在风景宜人的江景训练场内，多台形态各异的人形机器人正在学习端茶倒水、搬运重物等岗前技能。 上月底，湖北荆楚人形机器人有限公司（以下简称荆楚人形机器人）入驻武汉数创大厦。距大厦步行仅5分钟的一个新建1000平 方米生产工厂即将投入使用。 "这相当于楼上研发，楼下生产。"荆楚人形机器人总经理赵兴炜介绍，公司成立虽仅半年，但发展迅速，目前已研发出面向教 育、工业、家政和文旅等场景的四款人形机器人，签约订单金额已达数百万元。 在今年全省"新春 ...

Core Viewpoint - Cambricon announced a private placement with an issue price of 1195.02 CNY per share, raising a total of 3.985 billion CNY through the issuance of 3.3349 million shares, with GF Fund being the largest subscriber [2][5][6]. Group 1: Private Placement Details - The issue price of 1195.02 CNY per share represents a 110.51% premium over the minimum issue price [5]. - The private placement was conducted through a competitive bidding process, with the pricing benchmark set on the first day of the issuance period, September 18, 2025 [5]. - The total number of shares subscribed in this issuance is 3.3349 million, with a total subscription amount of 3.985 billion CNY [6][9]. Group 2: Major Subscribers - GF Fund received 1.0109 million shares, amounting to 1.208 billion CNY, ranking first among the subscribers [6][7]. - Other notable subscribers include UBS AG and Xinhua Asset Management, with allocations of 364,010 shares and 351,458 shares, respectively [7]. Group 3: Fund Utilization - The raised funds will be allocated to projects focused on chip platforms and software platforms for large models, as well as to supplement working capital [8][9]. - The investment amounts for the projects are as follows: 205.43 million CNY for the chip platform, 145.21 million CNY for the software platform, and 47.90 million CNY for working capital [9]. Group 4: Financial Performance - As of the end of Q2 2025, Cambricon reported total assets of 8.42 billion CNY and net assets of 6.76 billion CNY, reflecting increases of 25.34% and 24.58% year-on-year, respectively [10][11]. - For the first half of 2025, the company achieved a revenue of 2.88 billion CNY, a significant increase of 4347.82% compared to the same period last year, with a net profit of 1.04 billion CNY [10][11].

Core Insights - The 2025 World Manufacturing Conference was successfully held in Hefei, Anhui from September 20 to 23, focusing on the theme "Intelligent Manufacturing, Creating a Better Future" [13] - The conference attracted guests from over 40 countries and regions, featuring 4 major events, 6 project matchmaking activities, and 23 specialized activities [13] - The exhibition showcased achievements in new industrialization, particularly in artificial intelligence, with a dedicated robot section and the release of several large model technologies and application scenarios [13] Group 1 - The conference served as an important platform for international cooperation and was approved by the State Council, aligning with national diplomatic strategies [13] - A total of 735 cooperation projects were facilitated during the conference, with a total investment of 380.2 billion yuan, marking a 2.4% increase in project numbers and a 3% increase in investment compared to the previous year [15] - Manufacturing projects accounted for over 90% of the total, with the Yangtze River Delta region contributing more than 50% of the project numbers and total investment [15]

Core Viewpoint - The company, Keda Xunfei, has revised its fundraising plan for the issuance of A-shares, reallocating the use of raised funds to focus on computing power leasing in response to the rapid development of AI and large model technologies [1] Fundraising Plan - The total amount of funds to be raised remains unchanged, with the amount allocated for working capital reduced from 3.2 billion to 800 million [1] - 2.4 billion will be allocated for specialized expenditures related to computing power leasing [1] Computing Power Leasing Project - The company aims to implement a computing power platform project through a leasing model, expanding its computing power scale to support the development of large models and continuous algorithm innovation [1] - This project will enhance the base model upgrades for the Xunfei Spark cognitive large model and provide computing power support for the optimization of various models, including the Spark X1 model and multi-modal models [1] Strategic Implications - The expansion of computing resources is expected to improve the company's capability to provide multi-language large models for global clients, enhancing its technical strength for overseas AI business expansion [1] - The initiative is anticipated to create a new growth point for the company's business revenue [1]

Core Insights - The 10th Huawei Connect Conference (HUAWEI CONNECT 2025) was held in Shanghai, focusing on AI acceleration in financial technology (FinTech) innovation [1] - The chief algorithm scientist of Qifu Technology, Fei Haojun, highlighted three core challenges faced by microfinance: low data transparency, diverse enterprise types leading to disorganized data formats, and traditional risk control models lacking deep causal and logical reasoning capabilities [1] Company Developments - Qifu Technology is upgrading its core capabilities in perception, cognition, and logical decision-making to build a new generation of intelligent decision engines [1] - The company has developed a proprietary multimodal large model (MLLM) specifically for the financial sector, which excels in parsing vast amounts of non-standard documents like bank statements and income proofs, showing significant advantages in accuracy and recall compared to mainstream open-source models [1] - Qifu Technology has created three knowledge graphs covering 33 million enterprise entities, including enterprise relationship graphs, product graphs, and event graphs, enabling deep understanding of industry characteristics and complex relationships [1] Technological Innovations - The company has developed a reasoning large model (Reasoning LLM) for risk management in micro-enterprises, which features transparent interpretability and high efficiency, with traceable decision processes [2] - The model's accuracy and reliability are enhanced through long-chain reasoning data construction (LongCoT) and reinforcement learning (RL) mechanisms, allowing better understanding and explanation of micro-enterprise business operations [2] - Qifu Technology aims to enhance operational efficiency and risk control precision for financial institutions while promoting deeper understanding and precise services for each micro-enterprise [2]

Core Viewpoint - The 2025 Global Machine Learning Technology Conference will be held on October 16-17 in Beijing, focusing on cutting-edge AI research and applications, featuring over 50 prominent speakers from various fields [1][3]. Group 1: Conference Overview - The conference will cover twelve major topics, including advancements in large language models, intelligent agent engineering, multimodal models, and AI-enabled software development [3][4]. - The event aims to provide a platform for genuine exchange between academia and industry, showcasing both theoretical methodologies and practical experiences [4]. Group 2: Key Speakers and Sessions - Notable speakers include Lukasz Kaiser from OpenAI, Li Jianzhong from Singularity Intelligence Research Institute, and Wang Bin from Xiaomi Group, who will discuss the future of AI and large model technologies [6][14]. - The main stage will feature a high-level roundtable discussion on the core issues of AI industry paradigm shifts, involving key figures from the AI sector [14][15]. Group 3: Detailed Agenda - The first day will include sessions on topics such as the evolution of large language models and practical applications of multimodal models [15][28]. - The second day will focus on embodied intelligence, intelligent hardware, and the infrastructure needed for large models, with various specialized sessions scheduled throughout the day [22][28]. Group 4: Logistics and Participation - The conference will take place at the Westin Hotel in Beijing, with registration starting at 8:00 AM and the official program beginning at 9:00 AM on both days [31][32]. - Attendees are encouraged to arrive early to avoid congestion and ensure a smooth check-in process [32][33].

伴随AI普惠时代的到来，AI正在成为新时代创新生产力的"核心引擎"。早在2017年，联想就前瞻布局人 工智能，提出"联想 智慧中国"的愿景。如今，联想城市超级智能体基于先进的大模型技术和多智能体 协同架构，构建了业界领先的"1+N智能体"技术体系。该体系融合了深度学习、自然语言处理、知识图 谱、边缘计算等前沿AI技术，形成了具备自主感知、智能决策、协同执行能力的城市级智能系统。 本报讯 （记者贾丽）9月25日，在中国互联网协会主办的"2025数字政府智能应用与创新发展大会"上， 联想集团有限公司（以下简称"联想"）城市超级智能体凭借其创新的技术架构和丰富的落地实践，荣 获"2025数字政府创新解决方案"奖项，成为数字政府建设领域的标杆方案。 本次大会以"数智驱动 政务创新"为主题，汇聚了来自全国各地的政府代表、科研机构专家学者、行业 领军企业负责人。 业内人士认为，联想城市超级智能体能够在众多优秀解决方案中脱颖而出，充分体现了其在数字政府建 设领域的技术领先性和创新价值。这一荣誉不仅是对联想在人工智能、大模型技术应用方面创新成果的 高度认可，更是对其以技术驱动城市治理智能化理念的充分肯定。 在技术发展路径上， ...

在商家赋能层面，小智客服Agent打造场景化智能体，已覆盖商品咨询、国补咨询、商品比较、活动咨 询、尺码应答等二十多个高频场景，免配置、即开即用，同时商家自主可控可调优，无需额外运营投入 即可快速上线。文档知识问答能力更支持商家上传任意格式文档，系统会自动解析并转化为结构化知识 体系，大幅减少人工维护工作量，显著降低运营成本。 9月25日，2025京东全球科技探索者大会(JDD)在北京举行，大会正式发布京小智5.0——电商智能客服 行业Agent原生应用产品，该产品将于9月28日面向所有京东中小商家免费开放。 当前，电商行业智能客服已从"成本优化工具"加速向"品牌增长引擎"转型。作为京东百万商家的官方智 能客服，京小智5.0依托京东JoyAI大模型+多Agent协作技术架构，全面重构升级售前导购与全模态智能 客服能力，打造覆盖客服、导购、跟单、分析、质检的全链路Agent矩阵，将传统被动应答升级为具备 前瞻决策能力的主动服务。同时支持开箱即用的场景化智能体和自主可控的调优策略，真正让大模型智 能客服成为中小商家可轻松上手的"普惠应用"。 据5万家店铺内测数据显示，相较于升级前，京小智5.0帮助商家转人工率降低 ...